All techniques › Planting systems
Intercropping
Growing two grain crops in one field, usually a cereal and a legume. More grain per hectare at low input is well evidenced; drought insurance is not, and the hard part is what happens to the mixed grain at harvest.
The short answer
More grain per hectare at low input, yes. Across meta-analyses a cereal–legume mixture yields 1.2 to 1.3 times the average of its components grown alone, with higher cereal protein and fewer weeds. European wheat–pea, barley–pea and wheat–faba trials agree, most clearly on organic and low-nitrogen land.
Drought insurance, not shown. The only multi-site European stability trial, seven sites in 2017 and the drought year 2018, found intercrops more stable than sole crops only in more productive conditions, and as stable or less stable elsewhere. The buffering that exists is the cereal rescuing a failed legume. No trial shows a mixture using less water per hectare. That is why this page does not list the Mediterranean zone.
The harvest is the hard part. Two seed sizes in one combine, and a mixed lot that must be separated or sold as feed. Decide what happens to the grain before you sow.
The trade-off first
Intercropping costs almost nothing to try and can pay in the first season. What it costs is the harvest: both crops must ripen together, the combine must be set for two seed sizes, and the grain comes off mixed. Europe's largest intercropping project, ReMIX, with 23 partners in 13 countries, put it plainly: "separating grains after harvesting remains a major obstacle", its feasibility "depends on the species mixed" and on how much grain is broken, and "adopting intercropping does not depend on farmers alone" because buyers and value chains for mixed grain barely exist (ReMIX policy brief).
The farms for which none of that matters are the ones that feed the grain to their own livestock. For them the barrier disappears and the agronomy is all that is left.
What it is
Two grain crops sown and harvested together in one field, in a mixed stand or alternating rows. The European evidence is for cereal–legume pairs: winter or spring wheat with pea, barley with pea, wheat with faba bean, pea with oat. Two neighbours are dealt with on their own pages: trees with crops on the agroforestry page, and a companion that is never harvested on the cover crops page.
How it is meant to help
The two species use resources differently. The cereal takes soil nitrogen, which pushes the legume to fix more of its own; the cereal shades and props up the legume; the canopy closes earlier and suppresses weeds; and the cereal's grain protein rises because it competes less for mineral nitrogen (Bedoussac et al. 2015).
For drought the mechanism cuts both ways. An earlier, denser canopy reduces evaporation from the soil, and the two root systems can reach different water. But two species transpiring on one hectare can also use more water than one, and if the legume fails in a dry spring the cereal has to carry the field. What the European evidence actually shows is insurance against one component failing, not water saved.
What the evidence shows
Productivity: consistent, and mostly not European
| Synthesis | Size | Result |
|---|---|---|
| Li et al. 2023, PNAS (open access) | 934 observations, 226 experiments | Land equivalent ratio 1.23, a 19% land saving; 28% more yield than expected from sole crops; maize–legume mixtures gave 10% more protein per hectare than the best sole crop, and the advantage disappeared above 283 kg N/ha |
| Yu et al. 2015, Field Crops Research | global | LER 1.22 ± 0.02; the extra gain from staggered growing seasons appears in maize-based (C3/C4) mixtures, not in wheat–pea type (C3/C3) mixtures |
| Martin-Guay et al. 2018, Science of the Total Environment | global | LER 1.30; 38% more gross energy, 33% more gross income; as reported, paper not accessible to us |
Two things a European reader should weigh. A large share of the global data is Chinese maize systems, whose staggered-season advantage European cereal–legume pairs do not have. And the gains are largest at low input: a conventional cereal farm at high nitrogen rates should expect the bottom of the range.
Stability: the finding that decides this page's grade
Seven European field trials grew barley–pea and wheat–faba bean mixtures in 2017 and 2018, the second a drought year. Mean yields tended to be higher in the mixtures. But on stability the result contradicted the authors' own hypothesis: intercrops were as stable or less stable than sole crops at most sites, and more stable only where conditions were more productive; "yield stability significantly increased with increasing mean yields" (Weih et al. 2021, Agriculture).
For a site about drought that is the number one result. The stability benefit is weakest exactly where yields are already low and water-limited. The same shape as the cover-crop finding: a humid-climate benefit exported as if it were universal.
What buffering there is protects against the legume failing. In southern Sweden in the 2018 drought, faba bean grown alone yielded only 1.08 t/ha; in wheat–faba mixtures the loss was largely compensated by a larger wheat contribution (as reported; the primary paper was not retrieved). ReMIX's stability claim is likewise against the grain legume sole crop, not the cereal. When you read that intercrops are "more stable", ask: more stable than what.
Water: no saving shown
No European rainfed trial we found shows a cereal–legume mixture using less water per hectare than a sole crop. Egyptian deficit-irrigation trials report higher water-use efficiency per unit of yield (Frontiers, 2026), which is a different thing and under irrigation. A Swedish greenhouse study at 30% field capacity found intercropping "positive or neutral" for wheat and faba bean, with wheat's nitrogen uptake holding up better in the mixture (Frontiers in Agronomy, 2021), useful for direction and nothing more.
Other benefits: well supported
Higher cereal protein, weed suppression and reduced nitrogen fertiliser are consistent across European organic trials, and UK farmer field labs report 1.2 to 1.3 times the combined output per hectare, with final results still pending (Innovative Farmers).
Where it fits
| Context | Fit | Why |
|---|---|---|
| Organic or low-input arable above ~700 mm, grain fed on farm | Best | Largest gains at low input; no separation problem |
| Conventional arable at high nitrogen | Weak | Overyielding shrinks toward zero above ~283 kg N/ha |
| Malting barley or milling wheat | Avoid | Admixture breaks both; the protein rise breaks malting barley, where protein is capped, but helps milling wheat |
| Low-yield, water-limited arable | Not for drought | Stability gain absent in less productive conditions; use it for nitrogen and weeds if at all |
| Mediterranean rainfed grain | Not listed | No stability or water-use evidence found; the zone is excluded deliberately |
| Groves, vineyards, pasture | Not applicable | See inter-row cover and silvopasture instead |
What it costs
Almost nothing to sow; the cost sits at the other end of the season. No European per-tonne separation cost or mixed-grain price discount was found.
| Item | Figure | Country, year | Source |
|---|---|---|---|
| Separation by size, rotary screen cleaner | CAD 0.40–0.60 per bushel (roughly CAD 15–22 per tonne of wheat) | Canada, undated | No-Till Farmer, practitioner press |
| Separation by colour sorter | CAD 0.80–1.00 per bushel | Canada, undated | same |
| On-farm vibrating separator, up to 10 t/hr | Price not disclosed; payback "three to five years" | UK, 2024 | Farmers Weekly, reporting a vendor demonstration |
| Companion-cropping payment, Sustainable Farming Incentive | £55/ha | England, 2024 | same |
| Diversified crop structure practice | 233.13 zł/ha, about €55 | Poland, 2025 | funding page |
| Malting barley with pea | "not likely to be cost effective" | UK, older EU project | CORDIS |
What the schemes pay for. None pays for intercropping as such; they pay for the legume share or the diversity. In France, seed and forage legumes on at least 5% of arable land or more than 5 ha earn 2 écorégime points and 10% earn 3 (2025); a cereal–legume mixture counts in the legume category when it holds at least 50% cereal and 20% legume by usual seed rate. In Poland, the diversified-structure practice needs three crops with at least 20% of the area in soil-building crops, and mixtures qualify as long as cereals do not dominate (ARiMR). Rates and rules are on the funding page.
Timeline. Immediate. Benefits and costs both land in the sowing year; there is no transition dip and nothing to build up.
Commitment
Fully reversible, one season. No capital unless you buy a separator, and that is a machine, not a land commitment. The only lasting obligation is the year of any scheme declaration.
How to start
- Decide what happens to the grain first. Fed to your own stock, a mixture needs no separating and the main barrier vanishes. Sold, it needs a buyer or a cleaner arranged before sowing.
- Pick a proven pair with matched maturity: pea–oat or pea–barley in spring, wheat–faba or wheat–winter pea in autumn, in varieties that ripen together.
- Sow each at about half to two-thirds of its sole-crop rate, on one field, with a sole-crop strip of each component so you can work out your own land equivalent ratio.
- Set the combine for the larger seed and accept some loss of the smaller, or the reverse; test on a headland.
- Claim what the scheme pays: legume points in France, the diversified-structure practice in Poland, companion cropping in England.
- Do not expect it to fix a dry year. Expect it to stop a failed legume from emptying the field.
Risks and failure modes
Harvest mismatch: one component over-ripe while the other is green; legume lodging; shatter. The commonest reason farms stop after one season.
Separation and market: cleaning cost, a price discount, or no buyer at all.
Herbicides: what is safe for the cereal kills the legume. Intercropping is a mechanical or organic weed-control practice by default; its own weed suppression partly compensates.
No stability gain where it is most wanted. In the seven-site trial the mixtures were not more stable at the less productive sites. On water-limited land the two crops can fail together.
Water: no evidence of lower use per hectare, and some risk of higher use in a dry spring.
Specifications: mixed lots cannot go to a maltster or miller without cleaning.
Disease: a shared pathogen can move between components; reduced disease is the usual finding, not a guaranteed one.
Claims checked. Intercropping makes crops drought-resilient.
The seed order itself. Before mixing crops, the cheapest drought decision is which variety to sow; the drought-tolerant varieties page grades modern cultivars, landraces, sorghum and populations separately, and is the first technique here whose case is strongest in the Mediterranean.